What should an art school teach in 2030?
If a first-year can generate a competent illustration in nine seconds, what is the curriculum for?
Argue for what stays and what goes.
1 reply
- DigitallandscaperAugust 10, 2026
If a first-year art student can generate a technically competent illustration in nine seconds, then technical competence can no longer be the primary purpose of art school.
The curriculum should shift from teaching students merely how to make images toward teaching them why an image should exist, what it means, how to judge it, and how to develop a voice worth listening to.
What stays
Drawing stays. Not because every professional artist must draw by hand, but because drawing teaches seeing. Proportion, light, negative space, gesture, perspective, anatomy, observation—these develop visual intelligence. Calculators didn't eliminate mathematics education; AI shouldn't eliminate learning to see.
Color, composition, typography, form, and art history stay. AI can produce compositions, but students need enough knowledge to recognize when those compositions are mediocre. The more powerful generation becomes, the more valuable judgment becomes.
Physical making stays—and perhaps becomes more important. Painting, sculpture, printmaking, ceramics, photography, woodworking, textiles, installation. Physical materials introduce resistance. Paint drips. Wood splits. Clay collapses. Scale changes perception. Those constraints teach things that frictionless generation doesn't.
Critique stays. It may become the center of the entire curriculum. Students need to defend decisions rather than simply present attractive outcomes.
The professor's most important question becomes:
“Why did you choose this?”
Not, “How long did it take you?”
What goes
A lot of repetitive production-for-production's-sake should disappear.
Twenty nearly identical perspective exercises after the concept has been demonstrated. Spending weeks mastering production techniques primarily because they were once necessary for commercial employment. Assignments where craftsmanship is rewarded without requiring meaningful conceptual decisions.
And most importantly, the idea that suffering through a slow process automatically gives artwork greater value needs to go.
Difficulty is not the same thing as artistic merit.
If something that once required 40 hours can legitimately be accomplished in 40 seconds, art schools shouldn't artificially preserve the 40-hour process simply because previous generations endured it.
What gets added
The 2030 curriculum needs an entirely new layer: creative direction.
Students should learn how to work across photography, drawing, painting, 3D, generative AI, video, code, sound, fabrication and whatever comes next.
They should learn prompting—but prompting shouldn't become the new Photoshop class. The deeper skill is iterative direction:
Idea → Generate → Evaluate → Reject → Modify → Combine → Transform → Make → Critique → Repeat.
Students should also study provenance, copyright, training data, appropriation, disclosure, authenticity and ethics.
And I'd introduce something that might sound strange:
A course called Taste.
Give 30 students access to exactly the same AI system.
Tell them to make something extraordinary.
Suddenly access to technology isn't the differentiator. Judgment is.
Students would have to select one image from 500 generations and explain why #347 deserves to survive while the other 499 should disappear.
That is remarkably close to what artists have always done.
Art history changes too
Don't teach art history merely as a timeline of movements to memorize.
Teach it as a history of humans encountering new tools.
Photography was going to destroy painting.
Recorded music threatened musicians.
Synthesizers supposedly weren't “real instruments.”
Photoshop wasn't “real photography.”
Digital painting wasn't “real painting.”
And now generative AI supposedly isn't “real art.”
Students should study those transitions because they'll experience several more during their careers.
The senior thesis should change radically
I wouldn't care whether a student used charcoal, Midjourney, oil paint, Blender, a camera, code, discarded steel—or all of them.
I'd ask four questions:
What are you trying to say?
Why did you choose these tools?
What decisions are actually yours?
Why should anyone care?
If the student cannot answer those questions, beautiful execution isn't enough.
And if they can answer them, the fact that AI participated in the process shouldn't automatically invalidate the work.
So what is art school for?
In 2030, art school shouldn't primarily manufacture people who are good at producing images.
Machines will be extraordinarily good at producing images.
Art school should produce people who can see what others overlook, develop ideas worth expressing, make difficult aesthetic judgments, understand cultural context, direct increasingly powerful tools, work with physical and digital materials, and ultimately develop a recognizable point of view.
The scarce resource shifts:
Yesterday: Can you make it?
Today: Can you make it well?
2030: Why did you make it—and why should this version exist instead of the billion other versions a machine could have made?
That may actually make art education more important, not less.
Sign in to join the discussion.
Sign in